1 / 24

Ying Liu and Jay Verkuilen

Item Response Modeling of Presence-Severity Items: Application to Measurement of Patient-Reported Outcomes. Ying Liu and Jay Verkuilen. Outline. Introduction A framework for P-S items Nominal response model Example Comments. Introduction.

elysia
Download Presentation

Ying Liu and Jay Verkuilen

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Item Response Modeling of Presence-Severity Items: Application to Measurement of Patient-Reported Outcomes Ying Liu and Jay Verkuilen

  2. Outline • Introduction • A framework for P-S items • Nominal response model • Example • Comments

  3. Introduction • The Presence-Severity (P-S) format uses a compound item to assess an event, such as occurrence of a symptom. • P-S items come in two parts: • A filter, the presence part, is used to check whether the respondent experiences the particular event in question. • The Severity part is often about the frequency, density, severity, or impact of the event.

  4. Introduction • Example: • Presence • In the last week, did you have dry mouth? (yes/no) • Severity • If yes, how much did it bother you? (not at all/a little/somewhat/a lot)

  5. Introduction • What is a meaningful internal consistency reliability for the scale (or information curve)? • Are there items that do not behave appropriately given reasonable assumptions about the response process, in particular that lack of Presence should imply no Severity? • Is the format desirable or would it make sense to revise to a different format?

  6. Introduction • To combine the parts by appending the P part to be the bottom category of the S part, then a classical test theory reliability coefficient could be computed. • Only to analyze the P component by using a binary IRT model and ignoring the S part. • GRM or GPCM are considered to analyze P-S items. • Using Bock’s (1972) nominal response model (NRM) to analyze the P-S formats.

  7. Introduction • Investigating the necessity of the P-S format, especially in comparison of competing designs. • Examining the separation of the Presence and the Severity parts of the structure. • Using data from the Memorial Symptom Assessment Scale-Short Form (MSAS-SF), a widely used P-S format instrument to assess symptom distress in hospital patients.

  8. A Framework for P-S Items

  9. A Framework for P-S Items

  10. A Framework for P-S Items

  11. A Framework for P-S Items P=2 p20 p21

  12. Nominal Response Model (NRM) • a: category severity • c: relevant to the base rate of endorsing the category

  13. Nominal Response Model (NRM) • d10=a1-a0 reflects whether the P part is distinct from the bottom category of the S part. • If a1>a0, the P part is distinguished from the bottom category of the S part. • If a1<a0, the ordering required to append the filter item to the bottom of the S is not met.

  14. Example • Using MSAS-SF to assess symptom distress for cancer patients. • For each symptom, respondents are asked a “yes/no” question on whether they have had the symptom in the past 2 weeks. • Respondents who report to have the symptom are further asked to evaluate “how much did it bother or distress you” for the physical items and “how often did it occur” for psychological items.

  15. Example

  16. Example

  17. Example

  18. Example

  19. Example

  20. Example

  21. Comment & Question • It may generate a compound item with many categories. [14] • It was discussed that P-S format item may not be efficient and informative and thus should consider to abandon them. [14] • Is it more convenient to compose a questionnaire using compound items? • Will the use a1/ ao instead (a1-a2)more suitable parameter for judgment? [7]

  22. Comment & Question • Only a latent trait was involved in the present analysis. It is pity that the authors did not explain what is measured in the NRM. Obviously the nature of q is neither the health level (presence of physical/psychological symptoms) nor the frequency of falling ill (severity of physical/psychological symptoms). • We should try to use multidimentional or higher order IRT to trait this P-S format. I mean we can let P part as one latent trait and S part as another. [3]

  23. Comment & Question • Maybe the models used to fit the mixed-type items can be used in the future, such as the mixed-type model is the combination of the 3PLM (for dichotomous items) and the NRM or GPCM (for polytomous items).

  24. Thank you for your attention

More Related